Preparation of porous silica spheres for column liquid

Preparation of porous silica spheres for column liquid

Jowwal a$ Chronrato,grapfty, 83 (1973) 5-9 @ Elscvicr Scientific Publishing Company, CNROM. Amsterdam - Printed in The Nethcrlnnds 6805 PREPARATI...

336KB Sizes 0 Downloads 82 Views

Jowwal a$ Chronrato,grapfty, 83 (1973) 5-9 @ Elscvicr Scientific Publishing Company, CNROM.

Amsterdam

-

Printed in The Nethcrlnnds

6805

PREPARATION OF POROUS CHROMATOGRAPHY

SILICA

SPHERES

FOR

COLUMN

LIQUID

KLAUS UNGER and JURGEN SCHICK-KALB Tcclrrrisclrc HoclrscJrrrlc. 61 Darwrsladr (G.F. R.) and

KARL-FRlEDR1Ct-I KREk Fa. E. Merck. 61 Dartnstadf (G.F. R.)

SUMMARY

A new procedure for the preparation of porous silica spheres is described, which enables very small particles with a narrow size range to be prepared. Furthermore, the parameters of the pore structure can easily be controlled by the reaction conditions. In this way, porous silica supports of high efficiency were developed for the different types of liquid chromatography.

INTRODUCTlON

The development of supports of high efficiency is one of the most important problems in liquid chromatography. The column efficiency can be much improved by decreasing the particle size of the porous support, and as a result special types of porous silica have been developed with a mean particle diameter a,, < 1Opm (ref. 1). However, these supports should also have a definite pore structure, which depends on the type of liquid chromatography (partition, adsorption, ion-exchange or gel permeation) being considered. Firstly, the porous silica should be available with graduated values of the average pore diameter and the specific surface area. Supports for partition and gel permeation chromatography should exhibit a high specific pore volume at a homogeneous pore size distribution. On the other hand, the development of highly porous silica leads to the possibility of improving the column performance. As stated above, the diffusion path length of solute molecules can be shortened by using very small particles. However, the mean particle diameter of the porous support is limited to a value of about 5 pm, because it is very difficult to pack such columns with good reproducibility. In order to increase the accessibility of the pores of the small particles, supports should be synthesized with a high specific macropore volume. The macropores should be distributed on the surface of the particles as large holes and should have pore widths

6

K. UNGER,

J. SCHICK-KALB,

K.-F. KREBS

ten times smaller than the particle diameter (10004000 A). Furthermore, these macropores should also be the entrances to the smaller pores (intermediate pores). In order to attain all these requirements for a special pore structure, a new procedure of preparing porous silica was developed, which ‘was based on the earlier investigations of Kohlschtitter and Mihmz. The procedure and the dependence of pore structure on the reaction conditions will be published in detail elsewheres. This paper deals with the results, which are interesting for the development of silica supports of high efficiency for column liquid chromatography. EXPERIMENTAL

A general description of the preparation procedure is given below. The formation of porous silica consists of two stages. In the first stage, a viscous liquid polyethoxysiloxane is prepared by partial hydrolysis of tetraethoxysilane, and in the second step (Fig. I) the polyethoxysiloxane is emulsified with a mixture of water and ethanol. After emulsification by stirring, a catalyst is added, which initiates the hydrolytic polycondensation of the drops of polyethoxysiloxane to yield solid spheres of silica hydrogel. The hydrogel is separated from the solution and dehydrated to,yield porous silica. By varying the mean molecular weight of the polyethoxysiloxane and the reaction conditions (e.g., the type of catalyst, the catalyst concentration, the temperature, the relative proportions in the ethanol-water mixture) the parameters of the pore structure can be influenced and controlled.

Catalyst

I

Polyethoxysiloxane

bk

\ Emulsion

Fig. 1. Scheme of the procedure RESULTS AND

Suspension

for the preparation of porous silica spheres.

DISCUSSION

Particle size .and particle size range The particle size.range can easily be controlled by means of the stirring speed during the hydrolytic polycondensation of the polyethoxysiloxane. Fig. 2 shows tlie effect of the stirring speed on the particle size distribution. In this simple way, porous silica spheres with graduated average particle diameters and relatively narrow size ranges can be prepared. By subsequent sedimentation or sieving, very narrow cuts of particle size ranges are obtained. The spherical shape of the porous particles permits the reproducible packing, of the columns. .

PREPARATION

OF POROUS

SILICA SPHERES

1500

J

a z a!

FOR COLUMN

1300

1000

LC

750

rev. m in-l

6050. 40. JO20. lo5:

-

Mean particle

diameter

=p(urnm)

Fig. 2. Effect of the stirring speed on the particle size distribution. Average

pore

diameter

The average pore diameter of the porous silica spheres can be controlled in the range between 30 and 300 A by means of the catalyst concentration in the second stage of the preparation procedure. Table I shows the properties of a series of silica samples in which the catalyst concentration. was varied. With ,increasing catalyst concentration, the specific surface area decreases and the mean pore diameter increases. The products show a homogeneous pore size distribution. A maximum specific pore volume of 1.5 ml g-l was attained. These products were used as supports for adsorption and partition chromatography4. Supports with larger pores (mean pore diameter 15 > 300 A) were favourably prepared by the special treatment of the porous particles with a solution of NaCl and calcination of the salt-loaded gel at higher temperature#. Fig. 3 shows the integral pore size distribution curves of products that were treated by the “salt method”. The starting material for all samples was silica No. 5 in Table J. The average pore diameter can be varied in the range between 400 and 30,000 A (the maximum value TABLE I PROPERTIES OF POROUS SILICA SAMPLES WITH INTERMEDIATE SUPPORTS FOR ADSORPTION CHROMATOGRAPHY Specific srrrfaco area, S,,aT (,,t”g- ‘1

Spcclfic pore volume,

Meart port diameter,

VP (ml

D(A)

1

540

0.57

3

464

0.63 0.76

:o”

4 :

I%! 270 310

0.88 0.90 1.18

-: 140 80

NO.

6’

‘1

PORES

USED

AS

8 Cumulative

K. UNGER, pore volume

J. SCNICK-KALB,

K.-F. KREBS

v,,-,,

1000

loo00 Mean

Fig. 3. Integral pore size distribution intrusion technique.

pore

diameter

z(8)

curves of porous silica sphcrcs,

detcrmincd

by the mercury

of 30,000 A was obtained by using porous particles with an average particle diameter of about I OO,um). By varying the treatment conditions, not only the average pore diametkr but also the steepness of the integral pore size distribution can be controlled. Furthermore, the products exhibit a relatively large specific pore volume in the range between 0.7 and I .O ml g- I. These products are excellent supports for partition and gel permeation chromatography6. Specific pore volume

and porosity A special effect on pore structure is obtained by adding a solvent such as cyclohexane to the polyethoxysiloxane in the second step of the preparation procedure (Fig. 1) before starting the hydrolytic polycondensation. Because the formation of the spheres of the silica hydrogel occurs very rapidly, the added solvent is almost completely included in the solid particles. This mode of preparation yields a product with a higher specific pore volume than that without a solvent. Consequently, the specific pore volume of the porous spheres is a linear function of the amount of solvent added. In this way, we prepared samples with graduated values of the specific pore volume up to a maximum of about 4.0 ml g-l. From the rcRsults on the pore size distribution, we can conclude that the additional amount of the specific pore volume is due to larger intermediate pores (300 < 15 -=z2000 A) and to macropores (b > 2000 A). Two different specifications of the procedure are possible, yielding two kinds of products : (I) products with a nearly constant specific pore volume and graduated specific surface areas in the range between 200 and 500 mz g-l; (2) products with a nearly constant specific surface area and graduated specific pore volumes in the range between 1.0 and 4.0 ml g-l. Table II shows the properties of products that have nearly constant surface areas and graduated values of the specific pore volume. Sample 1 was obtained without adding cyclohexane. It has an average pore diameter of about 150 A and a homogeneous and symmetrical pore size distribution. Samples 2-5 were prepared by adding an increasing amount of cyclohexane. The samples show an unsymmetrical pore

PREPARATION TABLE

SILICA

‘SPNERES

FOR’COLUMN

LC

9

II

PROPERTJES NO.

OF POROUS

OF POROUS

SILICA SAMPLES

WITH GRADUATED

Specific surface area,

Specific pore voftmc,

Porosity

S,JST (+

VP (ml

P(%)

R-‘)

g-

‘)

1

270

0.97

66.0

f :

240 217 210 220

1.79 1.40 4.12 2.20

79.0 74.0 90.0 82.0

VALUES

OF POROSITY

-

size distribution. In the sequence of samples 2-5, the macropore volume increases, whereas the specific pore volume of the intermediate pores remains constant. The intermediate pores are mainly responsible for the value of the specific surface area and are accessible by the macropores. Using the products in Table II, we can investigate the effect of the accessibility of pores on the column efficiency in adsorption and partition chromatography. CONCLUSION

The objective of this paper has been to show the variability of the pore structure of porous silica spheres with respect to the development of supports of high efficiency for column liquid chromatography. REFERENCES .I. Chrotnatogw, Sci., 10 (1972) 593. H. W. Kohlschtitter and U. Mihm, Kollold-Z. Z. Polyrtr., 243 (1971) 148. K. Unger and J. Schick-Kalb, Kolloid-Z. Z. Pot’ytn.. submitted for publication, H. Engclhardt and N. Weigand, Paper presented at the I. Itrterrtarional Syrnposiutn Lkquid Chrotnatograplty, ftrtsrlakon, 2-4 May, 1973. 5 K.-F. Krcbs and H. Heinz, E, Merck, Darmstadt, Ger. Pat., 2.042.910. 6 K. Unger, unpublished results.

1 2 3 4

J. J. Kirkland,

ott Colwtrr